Combination MFB and Sallen Key Filter?

[T3sl4co1l]

I don't see that this topology exists anywhere, but it should work?

Prototypical circuit:



Assume zero-ohm sources.

Notice that MFB is inverting, while S-K is noninverting; both can be designed with gain.  Suppose we combine the circuits, for the same gain with respect to each input, zero common mode gain, and simplify.  We should get something like the above, but the trick is this:

What relationships hold for the various values of R and C?

R1/R5 = R3/R6 are fixed by the DC gain relationship.

That leaves R2, R4 and all the capacitors.

What other relations apply?

Related reading:
- I'm aware there's a fully-differential version; but without complementary outputs handy, what must the circuit look like?
- The 1st order version is quite simple: set R2 = R4 = C1 = C2 = 0, and Fc comes from the RC time constant (C4*R6 more or less).

Tim

[T3sl4co1l]

Playing around a bit (took some wrestling to make the PSPICE model work, as usual...   ), I see that R1 = R3, R5 = R6, R2 = R4, C1 = C2 and C3 = C4, as you would expect, to maintain good CMRR.  Mismatch leads to a bandpass / hump in the spectrum where poles and zeroes don't cancel out.

So, it'll have the same issue any other differential filter has, of course -- you need well matched capacitors and resistors to get good CMRR.

(The question that prompted this, has low CM anyway, so that's not a killer.)

Tim